[RavSinghChandan]

index2label casts the /Kids entry of a number tree to a list of dictionaries and
iterates it:

kids = cast(list[DictionaryObject], number_tree["/Kids"])
for kid in kids:
    limits = kid.get("/Limits", NullObject()).get_object()

The cast has no effect at runtime, so a file where /Kids holds a string raises
'AttributeError: str object has no attribute get' from reader.page_labels, and
a file whose kids array holds a number raises the same error for the element.

Both now take the fallback the function already uses for a /PageLabels entry it
cannot read: the malformed entry is reported and the page position is returned.
This follows #4022, which added the same handling one level up when /PageLabels
itself is not a dictionary.

The third test is a regression guard rather than a fix: an indirect kids array
and an indirect kid already resolved correctly, and .get_object() keeps that
working now that the cast is gone.

Offline suite passes, 1379 tests.

Used Claude (Opus) to help locate and patch this; I reviewed and tested the change.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 98.11%. Comparing base (095d5af) to head (e76448a).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #4102   +/-   ##
=======================================
  Coverage   98.11%   98.11%           
=======================================
  Files          59       59           
  Lines       11532    11539    +7     
  Branches     2152     2154    +2     
=======================================
+ Hits        11315    11322    +7     
  Misses        123      123           
  Partials       94       94           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.